Archive · 21 Jan 2026
AI Agents Pose Security Risks to Encrypted Apps
Meredith Whittaker, the president of the Signal Foundation, says that AI agents, which are like smart helpers that can do things like book appointments or shop for you, can be dangerous for apps like Signal that keep messages private.
These AI agents need to access a lot of personal data, like your contacts and credit card information, to work well.
If someone hacks into these AI agents, they could get access to all that private data.
Whittaker says this is like breaking a barrier that keeps your data safe.
Amazon is even suing another company, Perplexity AI, because they think their AI agent is causing privacy problems.
So, while AI agents can be helpful, they also make apps less secure.
AI agents pose security risks to encrypted apps like Signal, according to Meredith Whittaker, President of the Signal Foundation.
AI agents require extensive data access, including sensitive information like contacts and credit card data, to function effectively.
Whittaker warns that AI agents with root access permissions can be hijacked, compromising user data.
Amazon is suing Perplexity AI for allegedly introducing privacy vulnerabilities through its AI agent.
Signal is an encrypted messaging app popular among journalists and activists, known for its strong privacy features.
- Who
- Meredith Whittaker, President of the Signal Foundation
- What
- AI agents pose security risks to encrypted apps like Signal
- When
- Discussed during an interview at Davos, Switzerland on Tuesday
- Where
- Davos, Switzerland
- Why
- AI agents require extensive data access and operate with minimal human oversight, increasing the risk of data breaches
- Meredith Whittaker
- President of the Signal Foundation
- Signal Foundation
- Nonprofit behind the Signal encrypted messaging app
- AI Agents
- Autonomous systems that perform tasks like booking appointments or shopping
- Amazon vs. Perplexity AI
- Amazon suing Perplexity AI over alleged privacy vulnerabilities
- Signal App
- Encrypted messaging app popular among journalists and activists
- Data Access
- AI agents require access to sensitive data like contacts and credit card information
- Security Risk
- AI agents operate with minimal human oversight, increasing security risks
